I never lived the 12 Week Year in its pure form. I read the book, realized pretty quickly that on its own it wasn’t enough for me, and built my own variant right away. It’s been called the Phoenix Progression ever since.
No AI post today. This one is about the system I use to decide what I work on in the first place. If you’re mostly here for the evals and observability stuff: no worries, that’ll be back soon. If you stick around, a quick heads-up: in my everyday life I use the same vocabulary as engineering teams. Sprints and backlogs, postmortems, tech debt, refactorings that never happen. Except the tickets are my own.
The Phoenix Progression isn’t something I made up from scratch. The system combines a few building blocks that all exist already: the 12 Week Year by Brian Moran and Michael Lennington provides the sprint mechanics, Stephen Covey’s 7 Habits the layer above it with values and vision, and a cycles framework the micro layers below it, made of deep work blocks and short intervals. Wrapped around all of that is the phoenix arc, from burning through the ashes to the next attempt. That’s all the mysticism there is to the name.
What sets it apart from a plain 12 Week Year is the bracket above and below, plus a standalone ash week in between. But I’m getting ahead of myself.
Where the 12 Week Year fell short for me
The core idea from Moran and Lennington is strikingly simple. Instead of a twelve-month planning horizon, you take twelve weeks. “A year is no longer 12 months. It’s 12 weeks.”, the authors write on their site. Vague annual goals become quarterly goals with a real deadline, and resolutions become lead and lag metrics that you actually look at in a fixed weekly rhythm.
What convinced me most while reading was the logic of lead metrics. More on that below. In a few places, though, I noticed that I didn’t want to adopt the system one to one.
The bracket above is missing. After the twelve weeks comes the next twelve-week block, and what are they all building toward? The book has a vision phase for that, but in practice it tends to get skipped. That wasn’t enough for me. I needed something more stable than a quarterly plan, a five-year bracket with clear values and guiding goals that don’t get renegotiated every quarter. That’s where Covey comes in for me.
The break in between is too narrow. Moran and Lennington do include a short transition at the end of the twelve weeks, a kind of buffer for review and new planning. Organizationally, that makes sense. But it isn’t a real breather. In practice, that’s where I push the carry-overs into the next sprint and plan the next block, and suddenly the buffer week carries exactly as much output pressure as the eleven before it. It falls short of what sports science calls recovery. More on that below.
The granularity below is missing. The 12WY operates at the week level. But my actual working life happens between “this week” and “in two hours”. Which block on Thursday morning gets me closer to the quarterly goal, and which one is just motion? The book mostly leaves that open.
For these gaps I looked for building blocks that had already been tested instead of inventing my own. Covey’s 7 Habits, specifically “Begin with the End in Mind” and “Put First Things First”, provide the five-year bracket and the foundation of values. A cycles framework provides the micro layers of two-hour and 30-minute blocks. The ash time as a standalone recovery week is my own addition, inspired by sports science and Alex Soojung-Kim Pang’s idea of “deliberate rest”. The phoenix gives the whole thing its frame.
Six time layers - from the era down to the spark
This is what the finished hierarchy looks like:
Day to day, I mostly work on three of these layers: Flame, 1-Week Fire, Firestorm. The rest is the frame that holds those three up.
- Phoenix Era (5 years) - Long-term vision, core values, big guiding goals. Deliberately kept stable and not renegotiated every quarter. This is where the Covey part kicks in.
- Phoenix Cycle (6 months) - Two Firestorms plus two Ash Times. At most three main goals per cycle. The layer where I decide what gets priority at all over the next six months.
- Firestorm (12 weeks) - This is where the 12 Week Year lives on, practically unchanged. Three goals, lead and lag metrics, weekly review.
- 1-Week Fire (7 days) - What concretely moves forward this week? One clearly scoped sub-goal, lead metrics checked daily, an honest review on the weekend.
- Flame (2 hours) - A deep work block for one thing. The goal is set before it starts and reviewed at the end. No email, no Slack, no context switching.
- Spark (30 minutes) - A micro sprint inside a Flame. Four Sparks per Flame, with short breaks in between. Sparks only exist in the context of a Flame.
Each layer has its own ritual. A weekly review is something different from an era checkpoint, and it should stay that way. If you cram both into the same Sunday evening, you end up doing neither of them properly. The most interesting layer, by the way, isn’t the Firestorm. It’s the week in between, when nothing is deliberately on fire.
Fire and ash - why the Ash Time isn’t a buffer
The Ash Time is one week, always following a twelve-week Firestorm. No output pressure, no “but I really have to finish this” nagging at the back of my mind. Ideally a few days when I wake up without an alarm, linger a bit too long over breakfast, and the calendar is mostly empty.
It isn’t a vacation. On vacation I’m away; in the Ash Time I’m present without an assignment. I’m here, I’m awake, in a narrow sense I’m even still working. Just not on deliverables, but on whatever was lying underneath the last twelve weeks.
For me, this is the biggest difference from the plain 12 Week Year. Moran and Lennington planned for a short transition at the end of the twelve weeks, review and new planning. In practice, that’s where I push the carry-overs into the next block and set up the next three goals. I’ve done everything the book expects, and still haven’t recovered for a single hour.
Why the ashes have to stand on their own
For a long time I treated recovery itself as a nice-to-have. If the sprint phase went well, the week after was a bonus. If it went badly, it was earned, but somehow also suspicious. Both nonsense, both coming from the same unspoken assumption: that recovery is the opposite of work. It isn’t. It’s part of it.
Tudor Bompa formalized this in his Periodization (1963). Sports science has worked with cyclical load and cyclical recovery for decades, and the decisive effect, supercompensation, happens in the phase after training, not during the training itself. If you train without recovery, you get tired and eventually injured. You don’t get stronger.
Alex Soojung-Kim Pang carried that argument over from sports to knowledge work in Rest (2016). His term for it is “deliberate rest”: conscious, practiced recovery as a skill. In that logic, work and rest are partners in the same movement.
Loehr and Schwartz applied this to knowledge work in HBR back in 2001. High performance over long stretches only works as oscillation, not as constant strain. Ignore that and you don’t get more productive, you just sit at your desk longer.
For my taste, the plain 12 Week Year doesn’t do enough here. A buffer week is not an ash week. That’s why I added a standalone recovery week to the system.
What happens in the ash week
At its core, the week is held together by a few conceptual anchors, not by a weekly schedule.
A rebuilding setup. Recovery is active rebuilding, not standing still, and it needs its own infrastructure: sleep without an alarm, morning light, low-key exercise (a walk is enough), nature every day. None of this is new or secret. What’s new is that in the system it’s mandatory, not optional.
A reserved slot for deep reflection. Underneath every twelve-week sprint there’s a structural pattern I never see clearly while the sprint is running. Too close, too fast, too much output. In the next section I’ll pin it down as the Iceberg Flame. In the ash week I give myself the room to dig up one of these patterns. Not a hard block, more like: whiteboard on Tuesday morning, a walk in the afternoon, a few sentences in the notebook in the evening. Whiteboard again on Wednesday. If nothing comes, nothing comes.
A pre-mortem for the Ash Time itself. Before the week starts, I write down how it’s most likely to fail. Going in exhausted and just surviving the first few days. Squeezing the reflection into a hard block and strangling it. Pulling the quarterly planning forward too early and turning the Ash Time into sprint time in disguise. Once the likely breaking points are named in advance, they’re half defused. The pre-mortem costs ten minutes; skipping it costs the whole week.
The ashes belong to the sprint
The reflex is to see the Ash Time as a reward. Made it, now I’m allowed to rest. That’s the wrong geometry. The Ash Time is part of the same movement as the sprint before it: out of the fire into the ashes, out of the ashes into the next fire. Drop either phase and you end up with a straight line instead of a cycle, and straight lines eventually run out.
No ashes, no new phoenix. That isn’t meant poetically; it’s the structural claim of the system.
Lead metrics instead of a to-do list
In the fire I don’t need a plan, I need a speedometer. To-do lists are speedometers for standing still: they show what I’ve ticked off, but not whether I’m heading in the right direction. In the Firestorm, that’s not enough. On Friday I need to know whether the week held up, before the result becomes visible. That’s what Moran and Lennington’s split into lead and lag metrics is for.
Lag metrics are the results: revenue, weight, published posts, customer logos on the landing page. Honest numbers, but late ones.
Lead metrics are the upstream activities that I believe will lead to those results. Number of sales calls per week, completed training sessions, hours of writing per week, finished refactorings. The assumption I have to make on Sunday evening is: if I do these activities at this frequency, the result will come in time. If that assumption is wrong, I don’t have to wait until the end of the quarter to find out.
In the weekly rhythm of the Phoenix Progression, the lead metric gets scored, not the lag metric. If I already know on Friday that the week was a slog, it wasn’t the result that told me, it was the lead indicators.
What that looks like in the Phoenix Progression
For each Firestorm goal I define a handful of lead metrics, roughly three to five, depending on the goal. Every week, I put target next to actual. A fictional weekly table, let’s say from a freelance author who is also chasing clients alongside her book:
| Metric | Target (period) | Actual | Status |
|---|---|---|---|
| Weekly review done | 1 / week | 1 | OK |
| Outreach calls | ≥ 5 / week | 6 | OK |
| Manuscript pages | ≥ 8 / week | 8 | OK |
| Marketing hours | 3 / week | 1 | Weak |
| Planned Flames lit | 3 / week | 0 | Missed |
Execution score: 58 out of 100, in other words a sluggish week.
I read the score diagnostically. A 58 tells me that a good third of the planned activities fell through, and that it hit the Flames hardest, the ones that were supposed to happen this week. I can work with that on Monday. With “not done” on a list, I can’t.
The reverse case matters to me: a 100-percent week is suspicious. If you hit everything week after week, you set your targets too low. The score is supposed to stretch me, not pat me on the head. If it never hurts, the Firestorm was planned too small.
Why this beats a to-do list
Lead metrics can do a few things that plain lists simply can’t:
- They show direction, not just completion. “Done” tells me something is gone; it doesn’t tell me often enough whether it was the right thing.
- They’re early. I only see a lag metric like “articles published per quarter” at the end of the quarter. A lead metric like “hours of writing per week” shows up on Friday, and I can react on Monday.
- They force me to decide before the week what’s supposed to count. A to-do list can still be filled up on Tuesday, a lead metric can’t. It’s either locked in or it isn’t.
The last point is the painful one. I can run to-do lists reactively, lead metrics only prospectively. On Sunday evening I have to know what I’ll be measured against on Friday, and that isn’t pleasant. But it’s the part the whole system rests on.
What lead metrics can’t do: they won’t rescue a car that’s driving in the wrong direction. If the same lead metric gets missed three weeks in a row, “try harder” won’t help. At that point it’s a signal. Either the metric doesn’t fit, the goal doesn’t fit, or I don’t fit the plan right now. And this is where the Phoenix Progression shows its limits.
Where the system breaks
Not every week is a 90-out-of-100 week. Some are 58s. And some I look at in Sunday’s review and think: that wasn’t really a 12 Week Year week, it was a 12-day reality that called itself a week. The system can take that, but it comments on it. I would never have adopted a method in which a bad week just quietly fizzles out.
The Iceberg Flame
I call a task an Iceberg Flame when it looks like a small, quick block (an hour, maybe two), but a bigger structural pattern sits underneath it. On top, the task is visible and seems doable; underneath lies the reason it hasn’t been done for weeks.
Example, conceptually: a pipeline refactor that has been on the plan for five weeks. Every week it slides further back because something more urgent came up. The first time it’s bad luck, the second time a bad plan, the third time it’s a pattern. And the pattern rarely has anything to do with the refactor. It has to do with whether I still want this refactor at all, or whether it’s only on the list because it was a good idea at some point.
The Phoenix Progression forces me to name this explicitly on the third postponement. The weekly review then doesn’t say “postponed”, it says third postponement - iceberg suspected.
The postponement counter
Every postponed task gets a counter, a number next to the entry. On the third postponement there’s no fourth one; there’s a decision with a few possible outcomes:
- Drop it. The task was an idea, not a commitment.
- Reframe it. Maybe it was cut wrong and needs a different shape before it can even be started.
- Move it into the Ash Time. As an iceberg, not as a to-do.
That sounds fiddly, but it’s the mechanism that turns silent procrastination into a visible decision. As long as nobody’s counting, putting things off costs nothing. As soon as someone counts - in my case that’s the system itself - putting things off becomes a statement about my actual priorities. A task I’ve postponed three times isn’t “not done yet”. It’s been rejected; I just haven’t admitted it to myself yet.
What the Ash Time turns up
In the Ash Time I go through the icebergs one by one. Not to get them done (that would be sprint logic again), but to ask what they actually are.
What comes out of that is rarely an acute failure. It’s structural patterns:
- chronic procrastination in one particular category
- energy patterns without a clear trigger
- conflicts between two goals that block each other
- unresolved decisions that quietly hijack every weekly plan
At the end there’s a one-page diagnosis document with two questions, answered honestly: what showed up, and where does the system break. No to-do list. This document feeds the next sprint as a frame, not as a task list. It helps decide which Flames are even allowed into the next Firestorm.
Readable instead of successful
The system doesn’t make every week successful. It makes them readable. A single 58 week is information; three of them with the same pattern are a diagnosis. Without a diagnosis nothing changes structurally, and every week repeats the same quiet defeat. With a diagnosis I at least have a chance to address the pattern - or to consciously ignore it, which is also an answer.
What changed for me
What has shifted for me is less about the what and more about how I decide.
No more cold start on Monday. Monday used to be the day I decided what mattered this week. Now Monday is the day I execute what was already decided in Sunday’s weekly review. The actual planning happens before the sprint, not during it, and that saves the hours I used to lose to friction.
I can tell tasks apart by layer. When someone sends me a new idea on Friday, I no longer ask “am I doing this?”, but “which layer does this belong on?”. A Flame for next week, a weekly goal in the current sprint, a candidate for the next Firestorm, or an iceberg I won’t touch until the Ash Time. The answer decides effort and timing, not yes or no.
Recovery is planned in, not squeezed in. That vague feeling of “really needing a week off at some point” is gone. The Ash Time is in the calendar, four times a year, with dates. It’s just as much a part of the system as the sprint before it.
The long horizon is visible again. Before the Phoenix Era, 2030 was an abstract date somewhere beyond the next quarter. Now it’s the end point of the first era, with guiding goals and values that I recognize in my weekly plan. Not every week obviously pays into it, but I can tell which ones do and which ones don’t. That alone changes what I’m still spending my Wednesday evenings on.
Silent procrastination now has a price. What used to go unnoticed now costs a decision. That doesn’t always feel good, but it’s more honest than the old version, in which I told myself week after week that next week would be better.
What the Phoenix Progression did not do is turn me into a little productivity machine. There are still weeks when I write down a 58 on Sunday and wonder where the time went. The difference is that today I know where the missing 42 points are: in which Flame, in which iceberg, in which maintenance task I ignored for too long. That visibility is what was missing before.
I don’t think the Phoenix Progression fits anyone else in exactly the same way. Anyone who takes their time seriously ends up building their own Frankenstein system anyway. But here’s what I’m now fairly sure of: a sprint system without ashes you actually take seriously will sooner or later fall over exactly where the 12 Week Year fell over for me. Not in the sprint, in the gap after it.
Fire needs ashes to become new fire.
